McCabe PH, McNew CD, Michel NC. Effect of Divalproex-Lamotrigine Combination Therapy in Frontal Lobe Seizures. Arch Neurol. 2001;58(8):1264-1268. doi:10.1001/archneur.58.8.1264
Patients with intractable frontal lobe seizures represent a difficult subclass of patients with epilepsy. When medications fail, surgical outcomes typically have not been as successful as medial temporal lobe resections. The combination therapy of valproic acid (divalproex sodium) and lamotrigine has shown promising results in patients with uncontrolled seizures.
To determine outcome in patients with intractable frontal lobe seizures who were treated with the combination of divalproex and lamotrigine.
A nonrandomized, open-label, add-on trial.
Outpatients evaluated and treated at a tertiary care referral facility.
Twenty-one patients between 16 and 65 years old were studied. Patients were required to have failed at least 3 prior trials with antiepilepsy drugs. Criteria for frontal lobe onset included 1 or more of the following: frontal lesion on scan, positive ictal single-photon emission computed tomographic scan, symptoms consistent with frontal lobe onset, or an electroencephalogram (surface or invasive) consistent with frontal lobe onset.
Patients were treated with divalproex-lamotrigine combination therapy for 1 year.
Main Outcome Measures
The main outcome measured was seizure reduction. Safety and tolerability were also evaluated.
Four patients discontinued therapy. Ten of the remaining 17 became completely free of seizures. Two rashes occurred, but did not lead to discontinuation of therapy. The most common adverse events were tremor and weight gain.
Divalproex-lamotrigine combination therapy is a reasonable alternative in intractable frontal lobe epilepsy.
PATIENTS with seizures arising from the frontal lobe represent the second most common area of seizure onset in patients with partial seizures. It is estimated that 60% of partial seizures arise in the temporal lobe and 30% arise from the frontal lobe.1 When patients with partial seizures fail conventional antiepileptic drug (AED) therapies, surgical intervention may offer a chance at seizure freedom. However, surgical intervention with frontal lobe onset poses many problems, including a much larger area to sample, larger resections, and eloquent areas that must be avoided. In an initial study, we reported 24 (32%) of 85 patients becoming seizure free using the combination therapy of valproic acid (divalproex sodium [Depakote]) and lamotrigine in patients with intractable epilepsy.2 Of the patients with focal onset, 7 had medial temporal lobe onset and 6 met criteria for frontal lobe onset. A poor response was seen in patients with medial temporal lobe onset. None became seizure free and only 1 showed a 50% or greater reduction in seizures. In contrast, 3 of the 6 patients with frontal lobe onset became seizure free, and an additional 1 had a 50% or greater reduction in seizures. We decided to evaluate the outcome of frontal lobe seizures treated with divalproex-lamotrigine combination therapy in a separate, open-label, add-on trial.
Criteria for enrollment in the study included the following: (1) frontal lobe onset documented by invasive monitoring, (2) a frontal lesion on imaging studies consistent with clinical seizure, (3) a surface electroencephalogram consistent with frontal lobe onset, (4) clinical symptoms of seizure consistent with frontal lobe onset as described by Bancaud and Talairach,3 or (5) an ictal single-photon emission computed tomographic scan showing frontal lobe onset. Twenty-one patients met study inclusion criteria. Patients were then classified as having a high or moderate probability of frontal lobe onset. Patients were assigned to the category of high probability of frontal lobe onset if invasive data were positive or if 2 or more of the remaining criteria were met. Patients were assigned to the category of moderate probability if only 1 criterion was met.
Twelve of the 21 patients were initially receiving lamotrigine and 9 were initially receiving divalproex. Patients who were receiving lamotrigine as the initial drug had their lamotrigine dose decreased by 50% and therapy with divalproex sodium, 250 mg 3 times daily, was started. In 2 patients, lamotrigine was only decreased by 30% to 40% with divalproex sodium being started at 250 mg twice daily. Divalproex was then titrated upward to obtain a serum level between 80 and 120 µg/dL. The target serum lamotrigine level was between 6 and 14 µg/dL.
If divalproex was the initial AED, the starting dose of lamotrigine was dependent on whether an enzyme-inducing AED was also being used. If an enzyme-inducing agent was present, therapy with lamotrigine was started at 25 mg every other day for 2 weeks, followed by 25 mg/d for 2 weeks, and then increased by 25 mg every 2 weeks, typically to a dose of 200 mg/d. If an enzyme-inducing agent was not present, therapy with lamotrigine was started at 12.5 mg every other day for 2 weeks, followed by 25 mg every other day for 2 weeks. Further increases were then made as in patients with enzyme-inducing concomitant therapy. End doses were typically lower than with patients receiving an enzyme-inducing drug, usually between 100 and 200 mg/d.
Following evaluation and enrollment in the study, patients initially treated with lamotrigine were then seen at 3 months for the first follow-up visit. Patients initially treated with divalproex were seen at 4 months for the first follow-up visit because of the longer titration. Both groups were then seen every 3 months for the first year of treatment. Patients were given the option to continue combination therapy for long-term, follow-up treatment or to have their therapy changed. Serum AED levels, complete blood cell count with differential cell counts, and liver function tests were performed at 3-month, 6 month, and 1-year follow-up visits.
All patients were required to keep accurate records of seizures by diary. Adverse events were documented in the diary and were also inquired about at each visit. Patients were instructed to call at any sign of a rash, fever, or adenopathy.
Seizure outcome was documented at the 3-month (or 4-month), 6-month, 9-month, and 1-year visits. Patients were classified as being seizure free, having a 75% or greater reduction in seizures, having a 50% or greater reduction in seizures, or having no change. Adverse events were reported at each visit. Rash, tremor, weight gain, and hair loss were specifically inquired about at each visit.
One-year seizure outcomes are summarized in Table 1. Ten patients (47.6%) were free of all seizures at 1 year. All 10 were seizure free from the 3-month visit until the 1-year follow-up visit. Five patients were in the high probability group and 5 were in the moderate probability group. One additional patient (patient 5) was seizure free at 3 and 6 months, but discontinued therapy secondary to tremor and is excluded from the seizure-free data. Four patients (19%) had a 75% or greater reduction in seizures, 2 (10%) had a 50% or greater reduction in seizures, and 3 (14%) had no change at 1 year. Of the patients with no change, 1 was seizure free at the 3-month visit, but seizures returned to baseline values. Finally, 1 patient went from daily complex partial seizures to only having simple partial seizures.
Four patients discontinued combination therapy. Two patients discontinued therapy because of a lack of efficacy, 1 owing to adverse events and lack of efficacy, and 1 because of adverse events. Tremor was the most frequently reported symptom, occurring in 11 patients (52%). Weight gain was the second most common adverse event, being present in 8 patients (38%). Four patients reported alopecia. Neutropenia was noted in 1 patient. There was no decrease seen in the total leukocyte count, but a decrease in the percentage of neutrophils occurred. Thrombocytopenia was noted in 2 patients. Both neutropenia and thrombocytopenia were more likely with divalproex levels exceeding 100 µg/dL and both responded to a decrease in divalproex dosage. Rash occurred in only 2 patients, both of whom were able to continue receiving therapy after a mild dosage adjustment.
Partial seizures arising in the frontal lobe are thought to account for 30% of all partial seizures, being second only to temporal lobe onset.1 However, refractory frontal lobe seizures may pose a more significant problem, since they may not be as manageable by surgical techniques. Seizure-free rates following temporal lobectomy range from 50% to 70%,4 whereas in the past, most medical centers reported the seizure-free rate following frontal lobe resection to be only between 10% and 25%.5 Laskowitz et al6 reported 67% of the patients with frontal lobe epilepsy becoming seizure free following surgery; however, several patients required more than 1 surgical procedure. More successful outcome may have been related to improved imaging techniques. Better outcomes with frontal lobe surgery have typically been associated with the presence of a lesion.7 However, a recent report by Jobst et al8 reported 67% of patients undergoing frontal lobe resections to become seizure free, with outcome being better in nonlesional patients.
Because the success rate of surgery in patients with frontal lobe epilepsy may depend on whether a lesion is present, an alternative treatment that is less invasive would be preferred. Successful treatment with AEDs in this patient group has been suboptimal. Our data suggest that the combination therapy of divalproex and lamotrigine may be an option for patients with frontal lobe epilepsy before invasive procedures, especially in nonlesional cases. Rash was not a major adverse event with this combination therapy. Tremor, weight gain, and alopecia were exacerbated in these patients. Neutropenia is also a concern and needs to be monitored.
Our experience with this combination2,9 has shown that a high percentage of refractory patients become seizure free. Other authors have also reported a good outcome using the combination therapy of divalproex and lamotrigine.10- 12 To our knowledge, this is the first report of the combination therapy of divalproex and lamotrigine being useful in this specific patient population. The mechanism of this high response is unknown, but is thought to be pharmacodynamic. One may question if the high response rate is caused by polypharmacy. However, prior studies suggest a low response rate using multiple AEDs after a patient fails to respond to single-drug therapy.13- 15 It is estimated that only 0% to 11% of patients who fail the first AED will become seizure free with the addition of a second AED. In a study by Kwan and Brodie,16 23% of patients receiving 2 AEDs were reported to be seizure free after failing monotherapy and no patients taking 3 AEDs became seizure free. However, the specific combinations were not given. Since divalproex and lamotrigine are broad-spectrum AEDs, it may also be theorized that this could explain the high success rate. However, one would then expect to see similar findings with various combinations of any of the broad-spectrum AEDs such as divalproex, zonisamide, topiramate, or lamotrigine. We have not seen this thus far and have not seen any published results suggesting this, although a recent study by Stephen et al17 may suggest a favorable response when topiramate is combined with carbamazepine, lamotrigine, or phenytoin. In addition, if the high success was purely because of both agents being broad spectrum, it would not explain the difference in response between patients with frontal lobe onset and those with medial temporal lobe onset.
More than likely, the difference in response between these groups is related to anatomical, neurochemical, and/or neurophysiological differences in the 2 regions. One interesting anatomical difference that can be proposed is related to deeper structures, especially the thalamus, which is known to be involved in absence seizures18- 20 and may also play an important role in focal epilepsies. Several studies show a more critical role of thalamic involvement in seizure propagation in frontal lobe seizures compared with temporal lobe seizures,21- 24 although one study suggests early thalamic involvement in temporal lobe seizures.25 In addition, anatomical connections with the thalamus differ significantly between the frontal and temporal lobes.26,27 There is extensive innervation of the frontal lobe from multiple thalamic nuclei, including the anterior nuclei, mediodorsal nucleus, lateral dorsal nucleus, and the ventral anterior and ventral lateral nuclei. This innervation differs with specific regions of the frontal lobe and the connections are reciprocal in nature. The temporal lobe, in contrast, receives less innervation from the thalamus and this innervation is primarily from the pulvinar, intralaminar nuclei, midline nuclei, and anterior nuclei. In addition, one major projection from the temporal lobe to the mediodorsal nucleus of the thalamus is not reciprocated. Lesional and stimulation studies in humans and animals have shown the importance of the thalamus in seizures.28- 30 In one patient study, the effect of thalamic stimulation was reported to be more pronounced in generalized tonic-clonic seizures, typical absence seizures, and focal motor and secondary generalized seizures when compared with partial complex seizures, although detail of the area of origin of partial seizures was not given.30 Finally, the most common adverse effect of divalproex-lamotrigine combination therapy is tremor, which can also be modified by thalamic input.31 Therefore, one possible mechanism of action may be that the combination of divalproex-lamotrigine is modifying thalamic circuits.
One limitation of the study is the lack of invasive monitoring to confirm frontal lobe onset in patients. Only one patient had the procedure performed. Other patients, however, had features consistent with frontal lobe onset based on clinical features, electroencephalographic findings, imaging studies, or ictal single-photon emission computed tomographic scans. Therefore, one cannot rule out the possibility of onset in other areas, especially the temporal neocortex. A direct comparison between frontal and temporal neocortical seizures would be interesting; however, the temporal neocortical seizures are often difficult to identify without invasive monitoring and the patients will typically have surgery done at the time of monitoring. Several of our patients who have failed to respond to divalproex-lamotrigine combination therapy have gone on to have such monitoring, which has revealed temporal neocortical onset. Although this may suggest that this group is less likely to respond to the divalproex-lamotrigine combination therapy, this cannot be used as definitive evidence that patients with temporal neocortical onset will not respond to the combination. Also, formal quality-of-life studies were not performed. Of the 4 patients who discontinued combination therapy, only 2 did so because of adverse events. Although this suggests that the combination therapy was well tolerated, formal testing would have added more support to this perception.
As newer drugs for patients with epilepsy enter the market, it becomes more difficult to determine which AED to choose next, or which patients to refer to surgery. More data on patient outcomes to particular AEDs or AED combinations in specific epilepsy syndromes become more crucial. Any data that can be obtained may help with the development of management algorithms for epilepsy and, ultimately, lead to better care of patients.
Accepted for publication February 20, 2001.
Corresponding author: Paul H. McCabe, MD, Department of Medicine, Division of Neurology, Room C-6840, Milton S. Hershey Medical Center, 500 University Dr, Hershey, PA 17033 (e-mail: firstname.lastname@example.org).